Free SSI Science of Diving Specialty

[Steve_C]

Many shore dives are shallow to deep.

 

[dmaziuk]

So I find that these SSI warnings should be taken seriously, and discarding them as "errors", or "those guys do not understand how a modern computer works" is not a satisfying way of criticising these online lessons. They appear prudent to me, and prudent is OK.

If you state that an oversimplified model inside your computer does not exactly represent what happens in your body and therefore trust the computer but use common sense, that's perfectly fine.

If you say that theoretical nitrogen absorption is increased but residual nitrogen has "an effect" that computer does not accurately account for, you're smoking bad crack. The credibility of your subsequent "common sense" has gone out the window.

 

[saxman242]

If you say that theoretical nitrogen absorption is increased but residual nitrogen has "an effect" that computer does not accurately account for, you're smoking bad crack.

So this is the topic I was thinking of.

Let's go into pure speculation mode for a moment since I have no clue how the algorithms are being executed on the computer specifically.

[**SPECULATION**] The material seems to take the greatest pause with using the computer in a deep -> shallow -> deep situation. If we look at the simplest version of the computer programming I can come up with, it would take the current depth and the delta in the depth to see where you are and where you're going. If going down, add nitrogen, if going up, release nitrogen. Now, in a deep-> shallow -> deep situation, you could have some tissues on-gassing and others off-gassing, depending on loading before hand. Now, I imagine that some modern computers take this into account, but if you told me that earlier/more rudimentary computers took a more simplistic approach that didn't factor that in, I wouldn't be surprised [/**SPECULATION**]

This, of course, is separate from the discussion that the computer is just following a model based on observations and doesn't ACTUALLY measure nitrogen in your body, but that's a different topic.

 

[Angelo Farina]

If you state that an oversimplified model inside your computer does not exactly represent what happens in your body and therefore trust the computer but use common sense, that's perfectly fine.

If you say that theoretical nitrogen absorption is increased but residual nitrogen has "an effect" that computer does not accurately account for, you're smoking bad crack. The credibility of your subsequent "common sense" has gone out the window.

The computer does not account accurately of what happens in your body. It accounts just approximately. Then a "reasonable" safety factor is applied, which hopefully makes up for the inaccuracies of the model.
Sorry, but this is how engineering works. Exactness does not exist...

 

[dmaziuk]

It does still make the same statement about nitrogen loading when going from shallow to deep though.

I, unfortunately, don't have enough knowledge of exactly how things are calculated to technically refute it, although it goes against my understanding. It wouldn't, however, shock me if the very early computers did calculate in that way when going shallow to deep and that it's just a very dated statement. I would like to see that part updated.

The only thing I can think of if the part of bubble theory where "bubble crushing pressure" set in the previous part of the dive affect subsequent bubble growth, and "shallow crushing pressure" followed by a deep dive results in bigger bubbles on ascent.

Since that came out, there has been zero empirical evidence that reverse profiles or sawtooth profiles result in worse bubble scores. That's part of the reason bubble theory's been falling out of vogue in the last decade and everyone's been going back to simple stupid ZHL.

 

[dmaziuk]

Exactness does not exist...

Of course not. The point is that the model is statistically "safe enough", and safer than existing alternatives. Other than not diving.

 

[BlueTrin]

So this is the topic I was thinking of.

Let's go into pure speculation mode for a moment since I have no clue how the algorithms are being executed on the computer specifically.

[**SPECULATION**] The material seems to take the greatest pause with using the computer in a deep -> shallow -> deep situation. If we look at the simplest version of the computer programming I can come up with, it would take the current depth and the delta in the depth to see where you are and where you're going. If going down, add nitrogen, if going up, release nitrogen. Now, in a deep-> shallow -> deep situation, you could have some tissues on-gassing and others off-gassing, depending on loading before hand. Now, I imagine that some modern computers take this into account, but if you told me that earlier/more rudimentary computers took a more simplistic approach that didn't factor that in, I wouldn't be surprised [/**SPECULATION**]

This, of course, is separate from the discussion that the computer is just following a model based on observations and doesn't ACTUALLY measure nitrogen in your body, but that's a different topic.

I doubt they mean that the computer had faulty implementations, but rather than the models used were maybe not made to model this kind of behaviour ?

Maybe we should email SSI to ask clarification, if we are speculating it means that the material is not good enough.

 

[dmaziuk]

Now, in a deep-> shallow -> deep situation, you could have some tissues on-gassing and others off-gassing, depending on loading before hand. Now, I imagine that some modern computers take this into account, but if you told me that earlier/more rudimentary computers took a more simplistic approach that didn't factor that in, I wouldn't be surprised

I would: this is taken care of by "tissue compartments" and their "half-times". Those were proposed by John Scott Haldane in 1908. I'd be very surprised if any dive computer made after 1908 didn't take them into account.

 

[saxman242]

I would: this is taken care of by "tissue compartments" and their "half-times". Those were proposed by John Scott Haldane in 1908. I'd be very surprised if any dive computer made after 1908 didn't take them into account.

I'm sure it does for off-gassing, but it's questionable of what they do in sawtooth profiles. The tables use Haldane's models too, but that doesn't make them suitable for sawtoothing.

 

[saxman242]

I doubt they mean that the computer had faulty implementations, but rather than the models used were maybe not made to model this kind of behaviour ?

Well sure. I'm not saying that they may have been faulty, just that they may have been overly simplified in early implementations.